The present invention relates to a communication path switching device, a communication path switching method and a communication path switching program, and more specifically relates to a technology that can suppress occurrence of a communication breakdown in maintaining a network device present on a redundant path.
In recent years, cloud computing and cloud service have been penetrating into society. Movement of aggregation of various types of systems such as a corporate system into a cloud data center is accelerated, and the cloud data center is becoming greater.
The provision form of cloud data center includes a public cloud. The public cloud is characterized in that it is a multi-tenant type in which a system of one or more tenants is accommodated on one cloud system. In such a multi-tenant environment, since each tenant expands various services on the cloud system, it is impossible to stop the cloud system in terms of convenience of a cloud operator.
For example, when a switch is inevitably stopped in order to maintain a network device such as a switch or a router, the stop of the switch may cause the stop of the cloud system itself. Hence, the cloud operator previously plans to stop the cloud system and notifies a method of the stop and a period of the stop to all tenants who might be affected by the stop of the cloud system; depending on the situation, the cloud operator needs to receive permission to stop the system from each tenant. The cloud operator needs to perform such an operation for each maintenance, which places a significant burden on the cloud operator. Hence, it is required to provide a measure for maintaining the switch without the stop of the cloud system.
Hence, as a prior art, for example, a technology (see Patent Document 1, Japanese Patent Application Laid-Open Publication No. 2010-239279) is proposed in which two switches are coupled through two paths, an active path and a standby path, and when a failure occurs in the active path, the standby path is used instead of the active path to switch a communication path.
Moreover, a technology (see Patent Document 2, Japanese Patent Application Laid-Open Publication No. 2010-154383) or the like is proposed in which in a plurality of active paths coupling two switches (hereinafter, a plurality of paths coupling two switches is referred to as a redundant path), a path through the switch of a maintenance target is detected, a communication path for transmission to the switch of the maintenance target is switched to another path and thus the switch of the maintenance target is excluded from an arbitrary communication path.
When the prior art (for example, the technology of Patent Document 1) described above is used to stop the switch arranged on the active path and thereby switch the active path to the standby system, in the active path, both transmission processing and reception processing of communication might be stopped. Here, it is likely that it is impossible to receive, on the active path, communication which is supposed to be originally received by the switch, due to the stop of the reception processing of the switch. Therefore, the communication can be disadvantageously interrupted on a temporary basis.
In the case (for example, the technology of Patent Document 2) of a technology without consideration given to the transmission processing and reception processing of communication at the time of switching of paths, the communication is disadvantageously interrupted when both the transmission processing and reception processing of communication are stopped at the time of switching of paths.
On the other hand, as another method of excluding the switch of the maintenance target on the redundant path from an arbitrary communication path, for example, a technology is proposed in which a technology (for example, a technology disclosed in Non-Patent Document 1, IEEE Computer Society, “IEEE Standard for Local and metropolitan area networks-Link Aggregation”, IEEE [published on Nov. 3, 2008], [searched on Nov. 30, 2012], [Internet <URL: http://standards.ieee.org/getieee802/download/8020.1AX-2 008.pdf>) where in two switches sandwiching the redundant path, the transmission processing and reception processing of communication are independently controlled is utilized to stop only the transmission processing while the reception processing of communication is being operated on one path, and thus the communication with the one path is switched to the other path without interruption of the communication.
However, in the switch needed in the technology disclosed in Non-Patent Document 1, that is, the switch having the function capable of independently controlling the transmission processing and reception processing of communication, the cost of the device is more expensive than that of a normal switch. Hence, not all switches that perform the switching of the paths on a network necessarily have such a function; in actuality, switches having such a function and switches not having such a function are probably present in a mixed manner.
In a case where the communication path is switched in such an environment, when in order to switch the communication on one path to the other, the switch which does not have the function capable of independently controlling the transmission processing and reception processing previously stops both the transmission processing and reception processing of the communication on the one path, the communication is temporarily interrupted by the stop of the reception processing of communication.
In an actual network system, whether a redundant path is set at an active system or at a standby system may be changed depending on the transmission source of communication (see Non-Patent Document 2, VMware Inc., “vSphere Networking Guide”, VMware Inc., [date of issue is not disclosed] [searched on Feb. 19, 2013], [Internet <URL:http://pubs.vmware.com/vsphere-51/topic/com.vmware.ICbase/PDF/vsphere-esxi-vcenter-server-51-networking-guide.pdf>). In other words, in general, for the same redundant path, a plurality of different coupling policies such as only both the active systems, the active system and the standby system, and the standby system and the active system is set for each of the transmission sources of communication. Hence, it is easily expected that the prior art, in which the switching of the paths with consideration given to the coupling policies of the redundant path is not performed, cannot withstand the actual environment.